The purpose of the current research is to gather data on the effectiveness of a nationwide program aimed at reducing drinking/driving related problems, Students Against Driving Drunk (S.A.D.D.). S.A.D.D. is a multi-component intervention which includes a curriculum designed for high school sophomores, student committees aimed at altering school and community climate with regards to driving while intoxicated (DWI), and a Parent/Student Contract in which both students and their parents agree not to drive after drinking or to ride with an intoxicated individual. A number of social psychological mechanisms are proposed which may underlie the potential effects of S.A.D.D. Two interrelated studies are proposed. The first study is descriptive, and is intended to gather information concerning the operation of three well implemented S.A.D.D. programs in three Massachusetts high schools (S.A.D.D. originated in Massachusetts in 1981). In addition, the descriptive study will allow a pilot test of methods for the main, experimental research. The second study will be a true experiment comparing six schools in which S.A.D.D. is implemented with six matched control schools. Because pilot data suggest that many S.A.D.D. schools experience implementation problems, S.A.D.D. schools in the current research will receive a dissemination package developed by the research staff, and on-site technical assistance once the program is underway. Data collection for the research study will include significant attention to process issues, particularly program implementation. Assessments of program effectiveness will be made through comparisons of the S.A.D.D. and control schools on three school-wide student surveys and three random sample parent surveys (pre-implementation, post-implementation, and follow-up. a variety of multi-variate statistical methods will be used to assess program effectiveness and to isolate the underlying mechanisms contributing to program effects. The proposed program of research will result in a recommendation concerning whether the program (with modifications based on the research) is ready for wide-spread dissemination, whether significant modification and further study are needed prior to dissemination, or if the model is not viable in any form. If dissemination is justified, the dissemination package developed for the research will be revised to facilitate the development of new S.A.D.D. programs.